ABSTRACT TB treatment is an enigma of ineffectiveness. Current TB chemotherapy rapidly kills nearly all bacteria within two weeks, yet tolerable treatment failure rates are only achieved after 6 months, and even then, ~5% of cases relapse. Our recent work shows that bacterial factors associated with small MIC shifts are important predictors of treatment outcome, but the driving forces behind those shifts are unknown. This project unites three labs, with highly complementary expertise, around interrogating carefully curated M. tuberculosis clinical isolates with leading edge approaches in genetics, metabolism, gene regulation and network-based modeling to reveal fundamental new knowledge about how TB responds to front-line drugs. The direct result of this effort will be a suite of candidate biomarkers with great potential to personalize treatment duration by predicting treatment outcome and greatly simplify TB drug trials, as well as novel drug targets to improve outcomes and shorten therapy. These translational aims will be pursued in future studies, using the insights, strains and tools developed in the program.